The increase of manufacturing system flexibility makes variations in r
outings, operations, machines or operators possible. These degrees of
freedom may be used at the Production Activity Control (PAC) level in
order to react to unpredictable events such as machine failures, absen
ce of operators or changes in the workshop environment. In order to ma
nage these degrees of freedom, we suggest a Decision Support System th
at should make the choice of solutions easier in order to reduce the c
onsequences of a failure. This DSS complements the scheduling module o
f the PAC system since it may be used, either in order to slightly mod
ify an existing schedule, or to choose the scheduling hypothesis. This
DSS uses fuzzy logic and theory of possibility because of the impreci
sion and uncertainty of the information managed, first to model the ob
jectives, then to spread the expected consequences of an imprecisely k
nown event. The DSS has successively been prototyped with the SuperCar
d(TM) environment on a Macintosh microcomputer, then with the Oracle(T
M) data base on a Sun Sparcstation.